Traccar using 99+% of CPU (process & stack dump of process attached) making server Halt :(

Hamza Javed5 years ago

Operating System(OS): ubuntu
System spec. using lshw -short & lscpu: Image
OS Version Details :

  • Distributor ID: Ubuntu
  • Description: Ubuntu 16.04.6 LTS
  • Release: 16.04
  • Codename: xenial

Number of devices/trackers: 8
Server UP/Running time: Almost 9 days
CPU usage by traccar-server(JAVA-process): 98.00 - 99.97
PID of the top thread(JAVA) is: 21582 , HEX Value: 0x544E
In stack dump 0x544E refers to:

  • io.netty.channel.*
  • io.netty.util.*

To identify the thread followed steps in: Link
Image of my top using top -H: server-top
Complete stack dump of this process using jstack: thread-stack.txt

Any ideas whats the issue? why traccar using too much CPU and made server halt ?
any suggestions or information is appreciate. thanks
PS: if any other information required i'll be happy to share :)

Anton Tananaev5 years ago

Have you checked logs?

Hamza Javed5 years ago

yes apparently there isn't any error in logs except from these two

  • geocoder (Reverse-engineering) => 429 Too Many Requests
  • computed-attribute calculation => org.traccar.handler.ComputedAttributesHandler.computeAttribute@104![0,4]: 'adc0 * 5;' undefined variable adc0

but there is one thing more, take a look this might help threadDump.txt